WRONG QUESTION
“Every morning I check yesterday's marketing performance in our dashboards.”
Maybe you shouldn't be checking every morning at all.
The wrong question
Can it be automated?
Possible is not the same as sensible.
We can automate more than ever. That makes “can it be automated?” a surprisingly uninteresting question.
The harder question is whether it should be.
Some work is repetitive and predictable enough to hand to a system. Some work should disappear before anyone automates it. Some needs clearer structure first. AI can help with certain kinds of interpretation and variation. And some work is valuable precisely because a person is doing it.
We take the work apart first. Then we decide what could disappear, what could be simplified, what machines can do, where AI may help, and what is worth keeping human.
We take the work apart first
“Every morning I check yesterday's results in Google Analytics.”
What are you actually trying to achieve?
What we know
Checking Google Analytics every morning.
What it may mean
Knowing whether something important changed.
What work is hiding inside the task?
Does all of that work need to exist?
Maybe not. If the actual objective is knowing when something important changes, checking the same numbers every morning may not be necessary. A system could monitor continuously and ask for attention only when something deserves it.
Scheduled checking → exception-based attention.
What should do what?
Routine manual checking, if monitoring can replace it.
Retrieving and comparing known metrics.
Explaining or summarizing unusual changes where interpretation helps.
Understanding whether a change actually matters, and deciding what to do.
Anything that depends on information not provided here.
What might we be missing?
If looking through the numbers is a moment for reflection or spotting patterns nobody thought to ask about, removing the routine entirely may remove something valuable too. Reducing human effort is not automatically an improvement.
Then we answer
Partly.
Automate the checking. Keep the noticing human.
The archive
Things people have considered automating.
WRONG QUESTION
“Every morning I check yesterday's marketing performance in our dashboards.”
Maybe you shouldn't be checking every morning at all.
PARTLY
“Sales researches every inbound lead before contacting them.”
Automate the searching. Keep the understanding human.
YES, BUT…
“After every webinar we export attendees, clean the file and send the leads to Sales.”
Don't automate the handoff before asking why the handoff exists.
YES
“I copy new leads from one system into another and fix the fields before importing them.”
Moving and normalizing known data shouldn't need you.
What are you thinking about automating?